LeagueTLC Innovation Express
Exploring Issues, Innovations, and New Developments with Information Technology Professionals

Course-Based Model of Transfer Success: Rethinking Transfer Evaluation Michael Quanty  Senior Institutional Researcher Thomas Nelson Community College

American Community Colleges serve a legacy function in the preparation and transfer of students from 2-year institutions to the university. About a century ago, the first public junior college introduced itself into higher education. It was designed specifically to prepare students for a baccalaureate degree by offering the first two years and then having students transfer to a four-year college. Since that time, junior colleges have evolved into comprehensive community colleges, adding terminal certificates and associate degrees designed to prepare students for entry into the workforce as well as transfer. Similarly, former vocational and technical institutes have added college transfer curricula to their program mix. Although the evolution of the comprehensive community colleges has increased students' options in higher education, it also has complicated the evaluation of how well these colleges prepare students for transfer. By definition, community colleges respond to the needs of the local community. The transfer function may take a peripheral role compared to workforce training and economic development (Clowes and Levin, 1989). In fact, this summer Johnson County Community College and Oakton Community College hosted a conference entitled College Transfer: The Forgotten Function of Community Colleges. 

With research showing that nationally 42 percent of the students attending public community colleges intend to earn a bachelor's degree (Phillippe, 2000), why isn't transfer preparation a more prominent element in the community college's mission? The primary reason, according to some critics, is that community colleges do not do a good job of preparing students for transfer (e.g., Bernstein, 1986; Brint and Karabel, 1989). Carefully conducted research matching students entering community colleges with those entering four-year colleges shows that students who choose to begin at a four-year college are more likely to complete the bachelor's degree (Pascarella and Terenzini, 1991). Researchers over the years (e.g., Keeley and House, 1993) also have documented such phenomena as "transfer shock," a predictable drop in grade point average experienced by community college students during the first semester of transfer. While decades of transfer research have fueled the controversy about the community college's role in higher education, they have created no mandate for change nor widespread improvement efforts. The reason is simple-the typical research paradigm used to evaluate transfer does not provide information that colleges can use to actually improve the process. Institutions that are serious about articulation need data that can inform curricular improvement efforts.

Tracking Courses vs. Student Demographics

Over a decade ago, Thomas Nelson Community College (TNCC) and Christopher Newport University (CNU) decided that the key to improving transfer was to have teaching faculty involved from the outset in designing the research, reviewing results, and recommending changes based upon the research. With grant funding from the State Council of Higher Education for Virginia (SCHEV), four faculty members (two from each college) were enlisted to help design a comprehensive analysis of transfer. They were to assist research staff in developing a study to determine important factors that contribute to or inhibit successful transfer. They also were to conduct follow-up analyses, including hiring a consultant to interview a sample of students regarding their transfer experience. The faculty met weekly for a year. The research offices prepared detailed reports by pulling transcript information from both college databases on over 1,800 students. TNCC profiled successful and unsuccessful transfer students to examine differences between the groups using a variety of definitions of success (e.g., 2.0 GPA, % of successful grades, maintaining GPA), and performed discriminant and regression analyses to predict success and GPA. These various analyses produced a number of statistically significant trends based on such variables as hours completed prior to transfer, ethnicity, age, and gender. Some of the findings were surprising, some were disturbing, but none of them suggested corrective action. For example, the finding that older women performed relatively better than other groups produced a number of hypotheses and raised interesting questions. Nevertheless, the faculty members were at a loss as to how this finding could be used to improve the transfer prospects of other groups.

The frustration generated by the failure of carefully designed research to yield action-oriented results led us to re-examine our methodological assumptions. We realized that the focus of much transfer research was misplaced. Like virtually all researchers studying transfer, we had concentrated on students as the unit of analysis. Results documenting gender and ethnic differences do not help faculty prepare students more thoroughly. The information is interesting, but it does not create any incentive to act. It identifies problems, not solutions. Faculty cannot change students' demographic backgrounds or make them complete at least 30 hours before transferring. Even if they find that graduates in their curriculum experience difficulty when transferring, faculty need more specific information to correct the problem.

What is Course-based Tracking?

To bridge this gap between research and action, TNCC and CNU developed the Course-Based Model of Transfer Success (CBMTS). This model shifts the paradigm to focus on courses as the unit of analysis, rather than students. CBMTS provides the information and motivation faculty need to improve students' preparation. It shows faculty how students who complete their course perform in subsequent courses that require their course as a prerequisite. Their students' performance is compared to that of students who complete the prerequisite at the four-year college or at other institutions. This pinpoints for faculty exactly where students experience transfer difficulty. It also creates a sense of urgency. Faculty can readily agree that students who pass their course should be prepared for subsequent courses that require that course as a prerequisite.

CBMTS is a sophisticated tracking model that identifies every course at a four-year college having a prerequisite that can be met at the college or at a community college. It produces a grade distribution for each of those courses that compares performance for students who completed the prerequisite at the college with the performance of students who completed the prerequisite at any number of community colleges (or at other institutions, if specified). 

Course-Based Model of Transfer Success and Results

With support from a FIPSE grant, between 1996 and 1998,TNCC developed and tested a generic version of the program that allows other colleges to adopt CBMTS. The program has been written and tested using all 23 community colleges and six universities in Virginia. The community colleges ranged in size from Eastern Shore Community College with a fall enrollment of 662 to Northern Virginia Community College with 35,221 students on five campuses and their Extended Learning Institute. The six universities varied in size from Christopher Newport University with fall enrollment of 4,878 to Virginia Tech with 27,208. Three were urban universities and three were residential. Their admission criteria covered a wide range of selectivity.

Results from Virginia have been very encouraging and demonstrate CBMTS's ability to make the task of improving transfer a very manageable undertaking. First, the data clearly indicate that, in the vast majority of cases, community college courses provide comparable preparation. We examined grades in 1,273 courses at six universities, comparing 38,768 grades for students who completed prerequisites at one of Virginia's 23 community colleges to 183,365 grades for students who completed prerequisites at the university. Using productive grades (A, B, C, P) versus non-productive grades (D, F, W, U) in the first attempt at the course as the criterion, in only 5.8 percent of the comparisons did we find students who completed prerequisites at the university performing significantly better. Students completing prerequisites at a community college actually outperformed university prepared in 0.8 percent of the comparisons. 

Results such as these demonstrate the usefulness of the CBMTS model in informing policies and addressing general issues. They provide a quick and powerful refutation of anecdotal evidence often proffered to impugn the integrity of community college preparation. In the vast majority of cases, the higher education system was functioning as intended. 

The original FIPSE project also clearly demonstrates another major strength of the CBMTS approach: its ability to transform an intractable situation such as the "transfer problem" into a set of clearly delineated opportunities for two colleges to make changes that can improve students' outcomes. In Virginia, colleges used the idea of "critical comparisons" to identify problem areas that should be examined first. They defined these as instances which met the following criteria: a specific community college had at least five students who completed the prerequisite for the class and the Fischer Exact Test indicated that the percentage of those students receiving a nonproductive grade (defined as D, F, or W) was significantly higher (at p < .05 ) than that obtained by students who completed the prerequisite at the university.

Using this definition, ten of the 23 community colleges had no critical comparisons. Of the remaining 13, only the two largest community colleges had more than four critical comparisons. Since their data covered multiple campuses and several of the critical comparisons involved the same prerequisite course, the workload for any given campus was still very manageable. The "critical comparison" represents a high standard because it includes W grades as unsuccessful attempts. Summary information showed that community college students were more likely to receive a W grade than were students who did not attend community colleges. Including W grades as unproductive assumes that students often use that option to avoid a damaging grade. It also is reasonable to assume that often W grades reflect more on these students' life situations rather than reflecting an academic problem. These situations do not automatically change when the student transfers, especially when the transfer is as a commuter student at the local university.

Current Status of CBMTS

With our FIPSE grant we were able to demonstrate the viability of creating a generic version of the CBMTS software and to show the model's ability to provide data that lead to improvements. Through a series of conference presentations, workshops, journal articles, and personal contacts, we have generated a strong level of interest in the project. A paper describing the model was chosen best at the convention of the Southeastern Association for Community College Research and later presented in the distinguished paper session at the American Educational Research Association. Following a workshop at the Association for Institutional Research 38th Annual Forum, colleges in Oregon and Utah adopted the methodology. We also presented the model to groups of registrars and admissions professionals (a paper presentation at the Southern Association of College Registrars and Admissions Officers and the keynote address and a workshop at the New York State Transfer Articulation Association meeting). Assessment Update and the Journal of Community College Research and Practice published articles featuring the methodology and its advantages over traditional approaches. In Virginia, the State Council of Higher Education for Virginia (SCHEV) and the Virginia Community College System (VCCS) have agreed to coordinate and maintain a statewide reporting system and the State Committee on Transfer (SCT) has endorsed it as the model for evaluating transfer. Nationally, we have a group of colleges eager to adopt the methodology. In partnership with the League for Innovation in the Community College, we will be approaching foundations about the possibility of supporting a national dissemination.

If you are interested in more information or would like to be included in our dissemination efforts, contact:
Michael Quanty
Senior Institutional Researcher
Thomas Nelson Community College

References

Bernstein, A. (1986). The devaluation of transfer: Current explanations and possible causes. In L.S. Zwerling (Ed.), The community college and its critics (pp. 31-40). New Directions for Community Colleges, No. 54. San Francisco: Jossey-Bass.

Brint, S., & Karabel, J. (1989). The diverted dream: Community colleges and the promise of educational opportunity in America 1900-1985. New York: Oxford University Press.

Clowes, D. A., & Levin, B. H. (1989). Community, technical, and junior colleges: Are they leaving higher education? Journal of Higher Education, 60 (3), 349-355.

Doherty, F. Community College Transfer Performance at JMU. Office of Institutional Research, James Madison University.

Keeley, E. J., & House, D. J. (1993). Transfer shock revisited: A longitudinal study of transfer academic performance. Paper presented at the annual meeting of the Association for Institutional Research, Chicago, IL.

Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students: Findings and insights from twenty years of research. San Francisco: Jossey-Bass.

Phillippe, K. (Ed.). (2000). National profile of community colleges: Trends & statistics, 3rd Ed. Washington, DC: American Association of Community Colleges.

Quanty, M.B., Dixon, R., and Ridley,D. (March/April 1998). Community College Strategies: A New Paradigm for Evaluating Transfer Success. Assessment Update Vol. 10, Number 2.

Quanty, M.B., Dixon, R., and Ridley, D. (July/August 1999). The Course-Based Model of Transfer Success: An Action-Oriented Research Paradigm. Journal of Community College Research and Practice.

Townsend, B. (manuscript accepted for publication) Redefining the community college transfer mission. Community College Review.